Real-Time Reverse Transcription-PCR Detects KS1/4 mRNA in Mediastinal Lymph Nodes from Patients with Non-Small Cell Lung Cancer

Non-small cell lung cancer (NSCLC) is the most common cancer-related cause of death for both men and women in the US. Standard therapies for patients with NSCLC include surgery, chemotherapy, and radiation therapy, and the stage of disease dictates choice of therapy. The current staging system for lung cancer uses the American Joint Committee on Cancer TNM system, and its goal is to classify patients into groups based on the extent of disease. This system relies heavily on the pathologic evaluation of the primary tumor (T), regional nodes (N), and distant metastases (M). Patients in whom mediastinal lymph nodes (MLNs) are involved (N2 or N3) are classified with stage III disease (1) and are generally considered inoperable. The recent identification of genes overexpressed in lung cancer (2)(3)(4) combined with advances in real-time reverse transcription-PCR (RT-PCR) provide the opportunity to establish sensitive and specific ways to analyze MLNs. In addition, molecular biology approaches using real-time RT-PCR are well suited to the analysis of lymph node tissue procured through minimally invasive procedures such as endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA). This technique enables reliable biopsy of MLNs without the need for general anesthesia or surgery (5). Given the advantages of EUS-FNA, we investigated the possibility that metastatic disease could be reliably detected in MLNs of NSCLC patients by real-time RT-PCR. To define the ability of real-time RT-PCR to detect metastatic NSCLC in MLNs, we procured by EUS-FNA nine MLNs containing metastatic NSCLC (five adenocarcinomas, one large cell carcinoma, one squamous cell carcinoma, and two uncharacterized carcinomas). For negative controls, we collected 30 cervical lymph nodes obtained by surgical resection. Protocols for tissue procurement and patient consent governing all aspects of this study were reviewed and approved by the Medical University of South Carolina Institutional Review Board. For EUS-FNA, a fine-needle apparatus commercially produced for use with EUS (EUS N-1; Wilson Cook Co.) was advanced into a target lymph node under high-frequency (7.5 mHz) ultrasound guidance. An occluding stylet within the needle lumen was used to minimize contamination from pass-through structures and was removed once the needle was in position. As suction was applied with use of a syringe, the needle was moved back and forth within the lymph node for ∼2 min. This procedure typically retrieved a specimen of pure lymph node cells of ∼0.5–2 cm3. Material from the EUS-FNA was placed on multiple slides. One set of slides was air dried, stained with Diff Quik stain (Mercedes Medical), and interpreted in the endoscopy suite for specimen adequacy and for the presence or absence of metastatic NSCLC. Another set of slides was fixed in 950 mL/L alcohol and stained later with Papanicolaou stain. Criteria for metastatic carcinoma were the presence of one or more cohesive clusters of neoplastic cells with characteristic epithelial morphology and the presence of numerous lymphocytes in the background. Duplicate samples were placed on ice and taken immediately for RNA processing and real-time RT-PCR. For potential molecular markers of NSCLC metastatic disease, we tested the epithelial carcinoma-associated genes KS1/4 (6), lunx(7), [also known as palate, lung and nasal epithelium carcinoma associated (Plunc) gene (8)(9)], CEA, CK19, and muc1. β2-microglobulin was used as an internal reference control gene. The sequences for primers used in this study are listed in Table 1 . Primers for KS1/4 and lunx were designed using Primer Express Software (PE Biosystems). RNA isolation and real-time RT-PCR conditions were as described previously (10), with the exception that 0.1 U of UngErase enzyme and 0.25 U of AmpliTaq Gold were used per 10-μL reaction. The amplification efficiencies (AE) of KS1/4, lunx, and CEA, were 100%, 100%, and 98%, respectively; they were obtained by using the formula: AE = 101/m − 1 (11), where m is the slope of the line determined from linear regression analysis (Microsoft Excel software) of serial 10-fold dilutions of cDNA (data not shown) prepared from the lung cancer cell line A549. The amplification efficiencies of other genes used in this study were determined previously (10). Primer pairs/amplicons analyzed by real-time RT-PCR. Forward primer is shown as upper sequence in the respective primer pair. Reverse primer is the lower sequence. Predicted Tm determined using Primer Express software. Measured Tm determined from dissociation profile analysis (12). Sensitivity of respective gene for NSCLC detection in Cy+ samples. Specificity of gene with respect to negative control EUS-FNA samples. Primer pairs/amplicons analyzed by real-time RT-PCR. Forward primer is shown as upper sequence in the respective primer pair. Reverse primer is the lower sequence. Predicted Tm determined using Primer Express software. Measured Tm determined from dissociation profile analysis (12). Sensitivity of respective gene for NSCLC detection in Cy+ samples. Specificity of gene with respect to negative control EUS-FNA samples. For each NSCLC-associated gene, ΔCt values for cervical control lymph nodes and cytology-positive (Cy+) MLNs were obtained from triplicate reactions. The ΔCt value is the difference between the threshold cycle (Ct) for a NSCLC cancer-associated gene and that for a β2-microglobulin internal reference control gene. Relative gene expression of the samples was derived from real-time RT-PCR data using the equation: (1 + AE)ΔΔCt (10), where AE is the amplification efficiency of the gene of interest, and ΔΔCt is the difference between the mean ΔCt value in cervical control lymph nodes and the ΔCt value in the respective test sample. ΔCt threshold values for marker positivity were set at 3 SD away from the mean of the cervical control data set (Fig. 1 ). We observed that of the markers examined, KS1/4 had the highest sensitivity for detection of NSCLC metastatic disease (nine of nine samples; 100%). The sensitivities for the other genes are listed in Table 1 . There was no apparent difference in expression profiles according to tumor histology (e.g., squamous vs adenocarcinoma; data not shown). Lunx was overexpressed in seven of nine samples (Table 1 and Fig. 1 ); it was not overexpressed in one adenocarcinoma sample and in one uncharacterized NSCLC. Multimarker real-time RT-PCR analysis of NSCLC. Real-time PCR analyses of EUS-FNA specimens from 10 control MLNs (□), 30 cervical control lymph nodes (○), EUS-FNA specimens from 9 cytology-positive MLNs (•), and EUS-FNA specimens from 40 cytology-negative MLNs (▴) were performed as described in the text, using primer pairs for the indicated genes. Relative gene overexpression was determined from the equation (1 + AE)ΔΔCt. Threshold values for overexpression for each gene were calculated as described in the text and are depicted by the dash located on the right side of each data set. During the EUS-FNA procedure, a needle is passed through the esophagus into a MLN. Contamination from the esophagus and other pass-through structures is prevented or minimized by an occluding stylet within the needle lumen, which is removed after the needle is positioned within the lymph node. To verify that gene overexpression observed in the Cy+ samples was not attributable to an artifact associated with the EUS-FNA procedure itself, we determined gene expression for 10 negative control (subcarinal MLNs) EUS-FNA samples (Fig. 1 ). Control EUS-FNA samples were obtained from 10 consecutive patients who had no history of cancer and who were undergoing endoscopy for other indications (e.g., evaluation of benign pancreato-biliary disease). We observed that KS1/4 and lunx were not overexpressed in any of the 10 EUS-FNA negative-control samples [i.e., specificity, 10 of 10 (100%); Table 1 ], providing evidence that overexpression of these genes in the Cy+ samples was attributable to metastatic cancer. Regarding the other markers, we observed that CK19, CEA, and muc1 were overexpressed in three, two, and one EUS-FNA control nodes, respectively. These results suggest that overexpression of CK19, CEA, or muc1 can be associated with noncancer events. Hence, these genes may have limited utility as NSCLC molecular markers for samples obtained by EUS-FNA. Although our sample size was small, the results of our analyses of control and Cy+ EUS-FNA samples suggest that KS1/4 is the most informative molecular marker of NSCLC metastatic disease. To explore the possibility that real-time RT-PCR might be capable of detecting occult micrometastases, we analyzed 40 cytology-negative (Cy−) MLNs from 27 NSCLC patients who had no evidence of metastatic disease. Overexpression of the high-specificity markers KS1/4 and lunx was detected in 2 of 40 and 0 of 40 Cy− MLNs, respectively. Overexpression of the lower specificity markers muc1, CEA, and CK19 was detected in 2 of 40, 10 of 40, and 18 of 40 Cy− MLNs, respectively. These results suggest that real-time RT-PCR has the potential to detect occult NSCLC micrometastatic disease. In summary, we tested five molecular markers (lunx, KS1/4, CEA, CK19, and muc1) for use with real-time RT-PCR of EUS-FNA specimens and evaluated their sensitivity and specificity for the detection of NSCLC metastases. Overexpression of KS1/4 as well as at least one other marker gene was detected in all nine Cy+ specimens. The mean number of markers overexpressed was 3.8. Overexpression of lunx or KS1/4 was not detected in any of the control lymph nodes and thus had the highest specificity for NSCLC. These results demonstrate that real-time RT-PCR combined with EUS-FNA has potential value for staging patients with NSCLC. Furthermore, this is the first study to establish that KS1/4 is an informative molecular marker of metastatic NSCLC and that its apparent diagnostic accuracy is superior to other candidate markers. This work was supported by US Department of Defense Grant DOD N6311600MDM0U01-SP0007 (to M.M.) and National Cancer Institute/NIH Grant R21 CA97875-01 (to M.B.W.)